Statistics for Hackers

Abstraction: Replacing classical statistics jargon with computational simulation approaches

Key points:

• Thesis: statistics concepts (p-values, confidence intervals, distributions) can be replaced by simulation-based methods accessible to programmers
• Core techniques: sampling, shuffling, cross-validation, and Bayesian methods implemented as code loops
• "If you can write a for-loop, you can do statistical analysis"
• Presented at PyCon 2016 by Jake VanderPlas; video at youtube.com/watch?v=Iq9DzN6mvYA
• Earlier version presented at StitchFix, September 2015

Connections: Jake Vanderplas · Pycon · Computational Statistics · Bayesian Inference

Source: https://speakerdeck.com/jakevdp/statistics-for-hackers